ATI Radeon HD 2900 XT vs NVIDIA GeForce RTX 3060 3840SP
We compared two Desktop platform GPUs: 512MB VRAM Radeon HD 2900 XT and 6GB VRAM GeForce RTX 3060 3840SP to see which GPU has better performance in key specifications, benchmark tests, power consumption, etc.
Main Differences
NVIDIA GeForce RTX 3060 3840SP 's Advantages
Released 13 years and 8 months late
Boost Clock1852MHz
More VRAM (6GB vs 512GB)
Larger VRAM bandwidth (336.0GB/s vs 106.0GB/s)
3520 additional rendering cores
Lower TDP (185W vs 215W)
Score
Benchmark
FP32 (float)
Radeon HD 2900 XT
0.476 TFLOPS
GeForce RTX 3060 3840SP
+2887%
14.22 TFLOPS
Graphics Card
May 2007
Release Date
Jan 2021
Radeon R600
Generation
GeForce 30
Desktop
Type
Desktop
PCIe 1.0 x16
Bus Interface
PCIe 4.0 x16
Clock Speeds
-
Base Clock
1627 MHz
-
Boost Clock
1852 MHz
828 MHz
Memory Clock
1750 MHz
Memory
512MB
Memory Size
6GB
GDDR3
Memory Type
GDDR6
512bit
Memory Bus
192bit
106.0GB/s
Bandwidth
336.0GB/s
Render Config
4
Compute Units
-
-
SM Count
30
320
Shading Units
3840
16
TMUs
120
16
ROPs
48
-
Tensor Cores
120
-
RT Cores
30
-
L1 Cache
128 KB (per SM)
256 KB
L2 Cache
3 MB
-
-
-
Theoretical Performance
11.89 GPixel/s
Pixel Rate
88.90 GPixel/s
11.89 GTexel/s
Texture Rate
222.2 GTexel/s
-
FP16 (half)
14.22 TFLOPS
475.5 GFLOPS
FP32 (float)
14.22 TFLOPS
-
FP64 (double)
222.2 GFLOPS
Board Design
215W
TDP
185W
550 W
Suggested PSU
450 W
2x DVI
1x S-Video
Outputs
1x HDMI 2.1
3x DisplayPort 1.4a
1x 6-pin + 1x 8-pin
Power Connectors
1x 12-pin
Graphics Processor
R600
GPU Name
GA106
R600 XT
(215RGMDKA13FG)
GPU Variant
GA106-400-A1
TeraScale
Architecture
Ampere
TSMC
Foundry
Samsung
80 nm
Process Size
8 nm
0.72 billion
Transistors
12 billion
420 mm²
Die Size
276 mm²
Graphics Features
10.0 (10_0)
DirectX
12 Ultimate (12_2)
3.3 (full)
4.0 (partial)
OpenGL
4.6
N/A
OpenCL
3.0
N/A
Vulkan
1.3
-
CUDA
8.6
4.0
Shader Model
6.7
